1 Project Description
Tidy Tuesday has a weekly data project aimed at the R ecosystem. An emphasis is placed on understanding how to summarize and arrange data to make meaningful charts with ggplot2, tidyr, dplyr, and other tools in the tidyverse ecosystem.
2 Dataset
Data came from Wikipedia. The data includes information such as franchise name, revenue generated, year created, and owners/creators of the franchise.
3 Setup
3.1 Load Libraries
if (!require("pacman")) install.packages("pacman")
pacman::p_load("tidyverse","visdat","extrafont","ggpubr","png")
theme_set(theme_classic())3.2 Import Data
df <- read_csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2019/2019-07-02/media_franchises.csv")4 Exploratory Data Analysis
- There are 321 observations and 7 variables.
- Revenue is shown in ‘billions’ so the numbers seem small.
- Many of the variables are characters and appear to need cleaning.
- There are no missing values.
- 1996 brought in the most revenue.
- Pokémon, Hello Kitty, Winnie the Pooh, Mickey Mouse & Friends, Star Wars, and Anpanman have earned the most revenue.
- Merchandise, Licensing & Retail is the highest earning revenue category.
4.1 View Data
glimpse(df)## Observations: 321
## Variables: 7
## $ franchise <chr> "A Song of Ice and Fire / Game of Thrones", ...
## $ revenue_category <chr> "Book sales", "Box Office", "Home Video/Enter...
## $ revenue <dbl> 0.900, 0.001, 0.280, 4.000, 0.132, 0.760, 1.0...
## $ year_created <dbl> 1996, 1996, 1996, 1996, 1996, 1992, 1992, 199...
## $ original_media <chr> "Novel", "Novel", "Novel", "Novel", "Novel", ...
## $ creators <chr> "George R. R. Martin", "George R. R. Martin",...
## $ owners <chr> "Random House WarnerMedia (AT&T)", "Random Ho...head(df)## # A tibble: 6 x 7
## franchise revenue_category revenue year_created original_media creators
## <chr> <chr> <dbl> <dbl> <chr> <chr>
## 1 A Song o~ Book sales 0.9 1996 Novel George ~
## 2 A Song o~ Box Office 0.001 1996 Novel George ~
## 3 A Song o~ Home Video/Ente~ 0.28 1996 Novel George ~
## 4 A Song o~ TV 4 1996 Novel George ~
## 5 A Song o~ Video Games/Gam~ 0.132 1996 Novel George ~
## 6 Aladdin Box Office 0.76 1992 Animated film Walt Di~
## # ... with 1 more variable: owners <chr>summary(df)## franchise revenue_category revenue year_created
## Length:321 Length:321 Min. : 0.001 Min. :1924
## Class :character Class :character 1st Qu.: 0.211 1st Qu.:1977
## Mode :character Mode :character Median : 1.000 Median :1991
## Mean : 4.742 Mean :1986
## 3rd Qu.: 5.000 3rd Qu.:1999
## Max. :80.000 Max. :2013
## original_media creators owners
## Length:321 Length:321 Length:321
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
## sapply(df, function(x) n_distinct(x)) %>% sort()## revenue_category original_media year_created owners
## 8 18 52 78
## creators franchise revenue
## 101 103 174df %>% count(revenue_category, sort=TRUE)## # A tibble: 8 x 2
## revenue_category n
## <chr> <int>
## 1 Box Office 76
## 2 Merchandise, Licensing & Retail 73
## 3 Home Video/Entertainment 64
## 4 Video Games/Games 51
## 5 Comic or Manga 27
## 6 Music 12
## 7 TV 11
## 8 Book sales 7df %>% count(original_media, sort=TRUE)## # A tibble: 18 x 2
## original_media n
## <chr> <int>
## 1 Video game 64
## 2 Manga 63
## 3 Novel 33
## 4 Comic book 29
## 5 Animated film 27
## 6 Television series 19
## 7 Film 17
## 8 Animated series 16
## 9 Anime 13
## 10 Animated cartoon 8
## 11 Digital pet 8
## 12 Book 5
## 13 Visual novel 5
## 14 Greeting card 4
## 15 Cartoon character 3
## 16 Comic strip 3
## 17 Cartoon 2
## 18 Musical theatre 24.2 Missing Values
View missing values in more detail.
#Visualize missing values
vis_miss(df, sort_miss=TRUE)
4.3 Data Cleaning
Clean fields to make them easier to read/visualize.
df <- df %>% mutate(original_media = case_when(original_media %in% "Book" ~ "Novel",
original_media %in% "Cartoon character" ~ "Cartoon",
original_media %in% "Visual novel" ~ "Comic book",
original_media %in% "Animated cartoon" ~ "Animated series",
TRUE ~ original_media),
franchise = str_replace(franchise, "A Song of Ice and Fire \\/ ", ""),
franchise = str_replace(franchise, "Wizarding World \\/ ", ""),
franchise = str_replace(franchise, "Middle-earth \\/ ", ""),
revenue_category = str_replace(revenue_category, ", Licensing & Retail", ""),
revenue_category = str_replace(revenue_category, "\\/Entertainment", ""),
revenue_category = str_replace(revenue_category, "\\/Games", ""),
revenue_category = str_replace(revenue_category, " sales", ""),
revenue_category = str_replace(revenue_category, "Comic or Manga", "Comic\\/Manga"))4.4 Visualizations
df %>% ggplot(aes(year_created, revenue, fill=revenue_category))+
geom_col()+
scale_fill_viridis_d(option="E")+
labs(title="Franchise Revenue by Year", x="Year", y="Revenue (in billions)")
df %>% filter(year_created > 1970) %>%
ggplot(aes(year_created, revenue, fill=revenue_category))+
geom_col()+
scale_fill_viridis_d(option="E")+
labs(title="Franchise Revenue after 1970", x="Year", y="Revenue (in billions)")
df_revenue <- df %>% group_by(franchise) %>% summarize(total_revenue = sum(revenue)) %>% arrange(desc(total_revenue)) %>% ungroup()
df <- left_join(df,df_revenue)
df %>% filter(year_created == 1996) %>% mutate(franchise = fct_reorder(franchise, total_revenue)) %>%
ggplot(aes(franchise, revenue, fill=revenue_category))+
geom_col()+
scale_fill_viridis_d(option="E")+
coord_flip()+
labs(title="Franchise Revenue in 1996", y="Revenue (in billions)")+
theme(axis.title.y = element_blank())
df %>% filter(total_revenue > 50) %>% mutate(franchise = fct_reorder(franchise, total_revenue)) %>%
ggplot(aes(franchise, revenue, fill=revenue_category))+
geom_col()+
scale_fill_viridis_d(option="E")+
coord_flip()+
labs(title="Top Franchises' Revenue", y="Revenue (in billions)")+
theme(axis.title.y = element_blank())
df_revenue <- df %>% group_by(original_media) %>% summarize(media_revenue = sum(revenue)) %>% arrange(desc(media_revenue)) %>% ungroup()
df <- left_join(df,df_revenue)
df %>% mutate(original_media = fct_reorder(original_media, media_revenue)) %>%
ggplot(aes(original_media, revenue, fill=revenue_category))+
geom_col()+
scale_fill_viridis_d(option="E")+
coord_flip()+
labs(title="Revenue by Original Media Type", x="Original Media", y="Revenue (in billions)")
df_revenue <- df %>% group_by(revenue_category) %>% summarize(category_revenue = sum(revenue)) %>% arrange(desc(category_revenue)) %>% ungroup()
df <- left_join(df,df_revenue)
df %>% mutate(revenue_category = fct_reorder(revenue_category, category_revenue)) %>%
ggplot(aes(revenue_category, revenue, fill=original_media))+
geom_col()+
scale_fill_viridis_d(option="E")+
coord_flip()+
labs(title="Revenue by Category", x="Category", y="Revenue (in billions)")
5 Final Visualization
Image found at http://www.greeklibrary.org/wp-content/uploads/2019/05/books.jpg
img <- readPNG("books.png")
df %>% filter(original_media == "Novel") %>% mutate(franchise = fct_reorder(franchise, total_revenue)) %>%
ggplot(aes(franchise, revenue, fill=revenue_category))+
background_image(img)+
geom_col(color="gray", size=0.2)+
scale_fill_brewer(palette="Blues")+
scale_y_continuous(expand = c(0.01,0))+
coord_flip()+
labs(title="Franchised Novels' Revenue by Category", y="Revenue (in billions)", fill="Category")+
theme(plot.title = element_text(hjust=0.5, face="bold"),
text=element_text(family="Candara", size=13),
plot.background = element_blank(),
panel.background = element_blank(),
axis.title.y = element_blank(),
axis.text = element_text(face="bold"),
legend.background = element_blank(),
legend.text = element_text(size=8),
legend.title = element_text(size=10))
ggsave("media-franchises.png")